feat(transformer): Add babel-plugin-transform-dotall-regex

.gitignore

@@ -30,3 +30,6 @@ tasks/prettier_conformance/prettier/
 
 # Ignore the failures directory, which is used to store the results of the codegen coverage tests
 tasks/coverage/failures/
+
+# IntelliJ
+.idea/

Cargo.toml

@@ -111,6 +111,7 @@ project-root              = { version = "0.2.2" }
 quote                     = { version = "1.0.35" }
 rayon                     = { version = "1.9.0" }
 regex                     = { version = "1.10.3" }
+regex-syntax              = { version = "0.8.2" }
 rustc-hash                = { version = "1.1.0", default-features = false, features = ["std"] }
 ryu-js                    = { version = "1.0.1" }
 ropey                     = { version = "1.6.1" }

crates/oxc_transformer/Cargo.toml

@@ -30,6 +30,8 @@ rustc-hash      = { workspace = true }
 
 serde = { workspace = true, features = ["derive"] }
 phf   = { workspace = true, features = ["macros"] }
+regex-syntax = { workspace = true }
+lazy_static = { workspace = true }
 
 [dev-dependencies]
 oxc_parser  = { workspace = true }

crates/oxc_transformer/src/es2018/dotall_regex.rs

@@ -0,0 +1,68 @@
+use std::rc::Rc;
+use lazy_static::lazy_static;
+
+use oxc_ast::{ast::*, AstBuilder};
+use oxc_span::SPAN;
+use regex_syntax::ast::parse::Parser;
+use regex_syntax::ast::{Ast};
+
+use crate::options::{TransformOptions, TransformTarget};
+
+/// ES2024: Unicode Sets Regex
+///
+/// References:
+/// * <https://babel.dev/docs/babel-plugin-transform-dotall-regex>
+/// * <https://github.com/babel/babel/blob/main/packages/babel-plugin-transform-dotall-regex>
+pub struct DotallRegex<'a> {
+    ast: Rc<AstBuilder<'a>>,
+}
+
+impl<'a> DotallRegex<'a> {
+    pub fn new(ast: Rc<AstBuilder<'a>>, options: &TransformOptions) -> Option<Self> {
+        (options.target < TransformTarget::ES2018 || options.dotall_regex).then_some(Self { ast })
+    }
+
+    pub fn transform_expression(&self, expr: &mut Expression<'a>) {
+        let Expression::RegExpLiteral(literal) = expr else {
+            return;
+        };
+        if !literal.regex.flags.contains(RegExpFlags::S) {
+            return;
+        }
+        literal.regex.flags.remove(RegExpFlags::S);
+        let mut parser = Parser::new();
+        if let Ok(mut pattern) = parser.parse(literal.regex.pattern.as_str()) {
+            Self::transform_dot(&mut pattern);
+            let pattern = self.ast.new_str(pattern.to_string().as_str());
+            let regex = self.ast.reg_exp_literal(SPAN, pattern, literal.regex.flags);
+            *expr = self.ast.literal_regexp_expression(regex);
+        }
+    }
+
+    fn transform_dot(ast: &mut Ast) {
+        lazy_static! {
+            static ref DOTALL: Ast = {
+                let mut parser = Parser::new();
+                parser.parse("[\\s\\S]").unwrap()
+            };
+        }
+        match ast {
+            Ast::Dot(_) => {
+                *ast = DOTALL.clone();
+            }
+            Ast::Concat(ref mut concat) => {
+                concat.asts.iter_mut().for_each(Self::transform_dot);
+            }
+            Ast::Group(ref mut group) => {
+                Self::transform_dot(&mut group.ast);
+            }
+            Ast::Alternation(ref mut alternation) => {
+                alternation.asts.iter_mut().for_each(Self::transform_dot);
+            }
+            Ast::Repetition(ref mut repetition) => {
+                Self::transform_dot(&mut repetition.ast);
+            }
+            _ => {}
+        }
+    }
+}

crates/oxc_transformer/src/es2018/mod.rs

@@ -0,0 +1,3 @@
+mod dotall_regex;
+
+pub use dotall_regex::DotallRegex;

crates/oxc_transformer/src/lib.rs

@@ -10,6 +10,7 @@
 mod context;
 mod es2015;
 mod es2016;
+mod es2018;
 mod es2019;
 mod es2020;
 mod es2021;
@@ -38,6 +39,7 @@ use crate::{
     context::TransformerCtx,
     es2015::*,
     es2016::ExponentiationOperator,
+    es2018::DotallRegex,
     es2019::{JsonStrings, OptionalCatchBinding},
     es2020::NullishCoalescingOperator,
     es2021::LogicalAssignmentOperators,
@@ -74,6 +76,8 @@ pub struct Transformer<'a> {
     // es2019
     es2019_json_strings: Option<JsonStrings<'a>>,
     es2019_optional_catch_binding: Option<OptionalCatchBinding<'a>>,
+    // es2018
+    es2018_dotall_regex: Option<DotallRegex<'a>>,
     // es2016
     es2016_exponentiation_operator: Option<ExponentiationOperator<'a>>,
     // es2015
@@ -116,6 +120,8 @@ impl<'a> Transformer<'a> {
             // es2019
             es2019_json_strings: JsonStrings::new(Rc::clone(&ast), &options),
             es2019_optional_catch_binding: OptionalCatchBinding::new(Rc::clone(&ast), &options),
+            // es2018
+            es2018_dotall_regex: DotallRegex::new(Rc::clone(&ast), &options),
             // es2016
             es2016_exponentiation_operator: ExponentiationOperator::new(Rc::clone(&ast), ctx.clone(), &options),
             // es2015
@@ -224,6 +230,7 @@ impl<'a> VisitMut<'a> for Transformer<'a> {
     fn visit_expression(&mut self, expr: &mut Expression<'a>) {
         // self.typescript.as_mut().map(|t| t.transform_expression(expr));
         self.react_jsx.as_mut().map(|t| t.transform_expression(expr));
+        self.es2018_dotall_regex.as_mut().map(|t| t.transform_expression(expr));
         self.regexp_flags.as_mut().map(|t| t.transform_expression(expr));
 
         self.es2021_logical_assignment_operators.as_mut().map(|t| t.transform_expression(expr));

crates/oxc_transformer/src/options.rs

@@ -30,6 +30,7 @@ pub struct TransformOptions {
     pub function_name: bool,
     pub arrow_functions: Option<ArrowFunctionsOptions>,
     pub shorthand_properties: bool,
+    pub dotall_regex: bool,
     pub sticky_regex: bool,
     pub template_literals: bool,
     pub property_literals: bool,

tasks/transform_conformance/babel.snap.md

@@ -1,4 +1,4 @@
-Passed: 352/1369
+Passed: 354/1386
 
 # All Passed:
 * babel-plugin-transform-numeric-separator
@@ -695,9 +695,7 @@ Passed: 352/1369
 * regression/gh-7388/input.js
 * regression/gh-8323/input.js
 
-# babel-plugin-transform-dotall-regex (0/3)
-* dotall-regex/simple/input.js
-* dotall-regex/with-unicode-flag/input.js
+# babel-plugin-transform-dotall-regex (2/3)
 * dotall-regex/with-unicode-property-escape/input.js
 
 # babel-plugin-transform-async-to-generator (1/43)
@@ -832,7 +830,7 @@ Passed: 352/1369
 * general/function-duplicate-name/input.js
 * general/object/input.js
 
-# babel-plugin-transform-typescript (97/158)
+# babel-plugin-transform-typescript (97/175)
 * class/abstract-class-decorated-method/input.ts
 * class/abstract-class-decorated-parameter/input.ts
 * class/accessor-allowDeclareFields-false/input.ts
@@ -845,6 +843,23 @@ Passed: 352/1369
 * class/parameter-properties-with-parameters/input.ts
 * class/private-method-override-transform-private/input.ts
 * class/transform-properties-declare-wrong-order/input.ts
+* enum/boolean-value/input.ts
+* enum/const/input.ts
+* enum/constant-folding/input.ts
+* enum/enum-merging-inner-references/input.ts
+* enum/enum-merging-inner-references-shadow/input.ts
+* enum/export/input.ts
+* enum/inferred/input.ts
+* enum/inner-references/input.ts
+* enum/mix-references/input.ts
+* enum/non-foldable-constant/input.ts
+* enum/non-scoped/input.ts
+* enum/outer-references/input.ts
+* enum/scoped/input.ts
+* enum/string-value/input.ts
+* enum/string-value-template/input.ts
+* enum/string-values-computed/input.ts
+* enum/ts5.0-const-foldable/input.ts
 * exports/declared-types/input.ts
 * exports/export-const-enums/input.ts
 * exports/export-type-star-from/input.ts

tasks/transform_conformance/src/test_case.rs

@@ -119,6 +119,7 @@ pub trait TestCase {
                 .get_plugin("transform-exponentiation-operator")
                 .is_some(),
             shorthand_properties: options.get_plugin("transform-shorthand-properties").is_some(),
+            dotall_regex: options.get_plugin("transform-dotall-regex").is_some(),
             sticky_regex: options.get_plugin("transform-sticky-regex").is_some(),
             template_literals: options.get_plugin("transform-template-literals").is_some(),
             property_literals: options.get_plugin("transform-property-literals").is_some(),